Hello, although I'm researching in a different area, I was inspired by your article and want to try incorporating FLOPs into the calculation of the loss function. In your code, where exactly is the method for calculating FLOPs, as shown in the figure below?
